【python opencv3.3图像处理】——图像二值化、图像梯度、图像金字塔、Candy边缘提取
1.图像二值化全局阈值分割“cv.THRESH_TRIANGLE”只有单个波峰时效果较好,分割细胞图一般用"cv.THRESH_OTSU"“cv.THRESH_TRUNC”,截断,大于阈值设为阈值“THRESH_TOZERO”,小于阈值设为0cv.threshold(gray,0,255,cv.THRESH_BINARY|cv.THRESH_OTSU)#自动寻找阈值,"cv.THRESH_OTSU